ResearchBenchmarksPokerRules on PokerRules In-distribution (test)Follow97.9Task AccuracyDeepObjectLog79.523284.294189.06593.8359Jun 19, 2025Evaluation ResultsMethodMethodLinksTask AccuracyConcept AccuracyObject Count AccuracyDeepObjectLog2025.0697.93281.6NeSy2025.0680.2318.6380.2